Genetic and Molecular Characterization of the Caenorhabditis elegans Spermatogenesis - Defective Gene spe - 1 7 Steven

Two self-sterile mutations that define the spermatogenesis-defective gene spe-I 7 have been analyzed. These mutations affect unc-22 and fail to complement each other for both Unc-22 and spermatogenesis defects. Both of these mutations are deficiencies (hcDfl and hDfl3) that affect more than one transcription unit. Genomic DNA adjacent to and including the region deleted by the smaller deficiency (hcDfl) has been sequenced and four mRNAs (including unc-22) have been localized to this sequenced region. The three non unc-22 mRNAs are shown to be sex-specific: a 1.2-kb mRNA that can be detected in sperm-free hermaphrodites and 1.2and 0.56-kb mRNAs found in males. hDfl3 deletes at least 55 kb of chromosome ZV, including all of unc-22, both male-specific mRNAs and at least part of the female-specific mRNA. hcDf1, which is approximately 15.6 kb, deletes only the 5‘ end of unc-22 and the gene that encodes the 0.56-kb male-specific mRNA. The common defect that apparently accounts for the defective sperm in hcDfl and hDfl3 homozygotes is deletion of the spe-I 7 gene, which encodes the 0.56-kb mRNA. Strains carrying two copies of either deletion are selffertile when they are transgenic for any of four extrachromosomal array that include spe-I 7. We have sequenced two spe-17 cDNAs, and the deduced 142 amino acid protein sequence is highly charged and rich in serine and threonine, but shows no significant homology to any previously determined protein sequence. S PERMATOGENESIS in the nematode Caenorhabditis elegans is an attractive model system for studies of the genetic control of cellular differentiation. C. elegans is favorable for these studies because it is a self-fertile hermaphrodite in which internal fertilization is extraordinarily efficient, with virtually every sperm fertilizing an oocyte in young hermaphrodites (WARD and CARREL 1979). Mutant hermaphrodites containing defective sperm lay oocytes and are self-sterile; mating these mutants to wild-type males permits recovery of the mutation from resulting outcross progeny. This simple screen has identified >58 genes that appear to affect spermatogenesis (HIRSH and VANDERSLICE 1976; WARD and MIWA 1978; ARGON and WARD 1980; EDGAR 1982; BURKE 1983; SIGURDSON, SPANIER and HERMAN 1984; L’HERNAULT, SHAKES and WARD 1988; SHAKES 1988; J. VARKEY and S. WARD, unpublished data; S. L’HERNAULT, unpublished data). Much of spermatogenesis and how it is affected in various mutants can be studied both in vivo and in vitro, which permits direct microscopic inspection of this process as it is occurring. Wild-type C. elegans spermatogenesis has been studied in detail by both light and electron microscopy T h e sequence data presented in this article have been submitted to the ’ Present address: Division of Biomedical Sciences, Washington UniverEMBL/GenBank Data Libraries under the accession number 15322. sity, St. Louis, Missouri 631 10. Genetics 134: 769-780 uuly, 1993) (KLASS, WOLF and HIRSH 1976; WOLF, HIRSH and MCINTOSH 1978; WARD, ARGON and NELSON 1981). The 4N primary spermatocyte buds off the rachis (a central syncitial cytoplasmic core) after entering pachytene and undergoes the first meiotic division. Subsequent development of primary spermatocytes can occur in vitro in the absence of any exogenously supplied hormones or any type of accessory cell, such as the Sertoli cell that is required for mammalian spermatogenesis (reviewed by SKINNER et al. 1991). Secondary spermatocytes divide to form spermatids and discard components that are not required by the gamete into the residual body during this division. Resulting sessile apolar spermatids activate to form motile bipolar spermatozoa in vivo (WARD and CARREL 1979) and also can be activated in vitro (NELSON and WARD 1980; WARD, HOGAN and NELSON 1983; SHAKES and WARD 1989a). These spermatozoa crawl by directed membrane flow of a single pseudopod (ROBERTS and WARD 1982a,b) that is supported by a unique cytoskeleton (reviewed by HEATH 1992) and, unlike most sperm, lack both an acrosome and flagellum (WOLF, HIRSH and MCINTOSH 1978; WARD, ARGON and NELSON 1981). Spermatozoa must crawl to maintain their position within the hermaphrodite’s spermathecae where fertilization occurs because many are pushed into the uterus by passing oocytes (WARD and CARREL 1979). Asymmetric cytoplasmic partitioning places all ri770 S. W. L’Hernault, G. M. Benian and R. B. Emmons bosomes and virtually all of the actin, tubulin and myosin of the spermatocyte into the residual body during formation of haploid spermatids (NELSON, ROBERTS and WARD 1982; WARD 1986). Additionally, many sperm-specific components are segregated to the spermatid in unusual bipartite organelles called the fibrous body-membranous organelle (FB-MO) complexes (WOLF, HIRSH and McIntosh 1978; WARD, ARGON and NELSON 1981; ROBERTS, PAVALKO and WARD 1986), which are found in sperm of many nematode species (reviewed by FOOR 1983). Once the spermatid separates from the residual body, the FBs disassemble but the MOs remain intact and assume a position just below the plasma membrane. The MOs play a secretory role when they fuse with the plasma membrane during activation of a spermatid into a spermatozoon. spe-I 7 is one of several C . elegans spermatogenesisdefective ( spe ) genes that has FB-MO structural and/ or functional abnormalities (WARD, ARGON and NELSON 1981 ; SHAKES and WARD 1989b; L’HERNAULT and ARDUENCO 1992; VARKEY et al. 1993; J. VARKEY and S. WARD, unpublished data; S. L’HERNAULT, unpublished data). FB-MO defects in spe mutants are usually associated with other abnormalities in cellular morphogenesis or function. In the case of spe-17, mutant spermatids have ribosomes on membranes of their FB-MO complexes, but they also have an eccentrically placed nucleus and are only -2/3 the size of wild-type spermatids. Subsequently, many MOs fail to fuse with the plasma membrane as spe-17 spermatids differentiate into spermatozoa that have abnormally short pseudopods (SHAKES and WARD 1989b). Although aspects of the spe-I 7 mutant phenotype have been described previously (SHAKES and WARD 1989b), the genetic basis of this phenotype has not been described. The two chromosome IV mutations that cause the spe-I 7 phenotype are unusual in that they are tightly linked to a lesion in the unc-22 gene. Typical null unc-22 mutants twitch because of defects in body wall muscle (BRENNER 1974; WATERSTON, THOMSON and BRENNER 1980), but are fertile and contain normal-appearing sperm (NELSON, ROBERTS and WARD 1982; SHAKES and WARD 1989b). The two mutations that affect both unc-22 and spe-17 were found to be deletions, which were named hcDfl and hDfl3. We show that hDfl3 deletes >55 k b while hcDfl deletes approximately 15.6 kb of chromosome IV. The region affected by and adjacent to hcDfl was cloned and sequenced, and the pattern of transcription in this region was determined. Northern hybridization studies of this region revealed one 1.2-kb malespecific transcript encoded by a gene within unc-22, and a second 0.56-kb male-specific and a 1.2-kb female-specific transcript encoded by two genes upstream of the unc-22 transcription start. All of these genes are affected by the hDfl3 deletion, but only the gene encoding the 0.56-kb male-specific mRNA is deleted in hcDfl. The spe phenotype of both hcDfl and hDfl3 worms was corrected by transgenes containing genomic DNA that spanned the 0.56-kb malespecific transcript. Analysis of the spe-I7 gene and its cDNA sequence reveals that this gene encodes a small, novel protein that is highly charged and rich in serine and threonine, but has no significant homology to previously sequenced proteins. MATERIALS AND METHODS Strains, culture conditions and genetic nomenclature: C. elegans var. Bristol (strain N2) was the wild-type strain used in all experiments (BRENNER 1974). The following genes and mutations were used in this study, LGlV, dpyl?(e184), dpy-ZO(e1282), unc-?O(e191), unc-22(ct?7, e2182) (MOERMAN et al. 1988; A. Z. FIRE, unpublished data),feml(hcl7ts) (NELSON, LEW and WARD 1978), fem-?(q2?& (BARTON, SCHEDL and KIMBLE 1987); LGV him-5(e1490) (HODGKIN, HORVITZ and BRENNER 1979). The chromosome I V deficiencies sDF8, sDP, sDfl9, (MOERMAN and BAILLIE 1981; ROGALSKI, MOERMAN and BAILLIE 1982; ROCALSKI and BAILLIE 1985) and SOP? u. E. SCHEIN, M. A. MARRA, G. M. BENIAN and D. L. BAILLIE, in preparation) and the translocation nT1 also was employed (FERGUSON and HORVITZ 1985). Previously, it was shown that the deficiency sDf83 failed to complement hDfl? for sterility u. E. SCHEIN and D. L. BAILLIE, unpublished observations). Culturing, handling and genetic manipulation of C. elegans were performed as described (BRENNER 1974), and standard nomenclature was used (HORVITZ et al. 1979). Origin of spe-17 mutations: The hcDfl deletion was selected as a spontaneous unc-22 sterile mutant with the strain name TR708, and was the generous gift ofJ. COLLINS and P. ANDERSON. hcDfl arose spontaneously in the Bristol/ Bergerac hybrid strain TR644 (COLLINS, SAARI and ANDERSON 1987). The hDfl? deletion was originally selected as a spontaneous unc-22(h12) Bristol mutant that was sterile; its strain name is KR127 and it was isolated in the laboratory of A. ROSE. This mutation was balanced to nT1 to make strain BC3 12 1 unc-22(h12)/nTi IV; n T l / + V this strain was the generous gift of D. CLARK and D. BAILLIE. The name h12 was used previously (SHAKES and WARD 1989b), but it was changed to hDfl? after the present work revealed that it was a deletion mutation. Worm culture: Routine worm growth was performed at 20” in agar filled petri plates that were seeded with Escherichia coli OP50 (BRENNER 1974). DNA was prepared from worm stocks raised on 9-cm agarose plates seeded with E. coli P9OC (MILLER et al. 1977; FIRE 1986). RNA was prepared from worms raised in liquid cultures that were synchronized by hypochlorite treatment (NELSON, ROBERTS and WARD 1982). Sex determination mutants were used to produce animals that had a female soma and either oocytes, but no sperm (fem-1; NELSON, LEW and WARD 1978) or sperm, but no oocytes (fern-3&, BARTON, SCHEDL and KIMBLE 1987). Males were prepared from him-5 cultures as described previously (KLASS and HIRSH 1981; NELSON, ROBERTS and WARD 1982). Nucleic acid methods: DNA was prepared from mixed stages of worms grown on plates as described, with minor modifications (SULSTON and HODCKIN 1988). Total and poly(A+) RNA were prepared as previously described (BURKE and WARD 1983; ROSENQUIST and KIMBLE 1988). Plasmid DNA was prepared for worm microinjection experiments by chromatography on Qiagen columns according to C. elegans spe-I7 Gene Analysis 77 1